introducing Unclutter: a smart ad + slop blocker with Jev 🤓 it auto cleans up pages from slop elements: ⬖ ads ⬖ cookie banners ⬖ upsells ⬖ bs dialogs BYOK. open source + free, download below 👇
Evaluation & Training
Cost/quality frontier
Plotting accuracy against cost and time on log scales, where up and to the left is better; cascades are designed to sit there.
- Category
- Evaluation & Training
- Also known as
- —
- Related terms
- 4
- Directory entries
- 7
- Docs
- evals.typesafe.ai
- Added
- 2026-09-24
Definition
The workflow evals chart accuracy against cost and time on log scales with up-and-to-the-left marked better. The SDE cascade cookbook uses the same framing, reporting that on an internal sweep the cascade sat up-and-left of every single model.
The frontier is a decision tool: a cascade or a Jev workflow is worth its complexity only if it moves the operating point, which is why the cookbooks publish both quality and price next to each other.
Tagsevaluationcost
Related terms
Definitions that connect to this one.
SDE cascadeA two-stage extraction cascade: a mini model extracts, a Noul battery verifies, and a confidence gate escalates to a reasoning model.PatternsWorkflow evalsTypeSafe's published evaluations of four automation workflows, comparing Jev and frontier LLMs as structured workflows versus single prompts.Evaluation & TrainingBenchmark claimsTypeSafe's headline multipliers — 193.6x faster, 444.6x cheaper, a 238x lower input price — come from its own workflow comparisons.Evaluation & TrainingToken pricingJev's published rate: $42 per billion input tokens, with output tokens free.API & Models
From the directory
Published evaluations of four automation workflows, security incidents, agent trace observability, invoice processing, and customer service, comparing Jev and frontier LLMs as structured workflows versus single prompts.
Practices & PatternsDocs#official#benchmarks#evaluation
Official
A two-stage extraction cascade: a mini model extracts, a Noul battery verifies each field in one request, and a 0.7 gate escalates to a reasoning model, sitting on the cost/quality frontier.
Cookbooks & DemosDocs#official#cookbook#extraction
Official
A natural-language smart-room demo: describe a mood and Jev decides lights, curtains, TV, music, volume, and desk lamp in one request.
Cookbooks & DemosPlayground#community#playground#smart-home
Community
Cole Medin's week of hands-on testing, covering where Jev gates replace regex or slow LLM calls, plus demos that play his indie game and triage every open pull request.
Sites & GuidesVideo#community#video#tutorial
Community
codebasics explains why a non-generative model matters and walks through hands-on code, including the caveat that Jev cannot write text or code.
Sites & GuidesVideo#community#video#tutorial
Community
A head-to-head benchmark that prices 200 products from descriptions with Jev against GPT-5.6 Luna and GPT-4.1-Nano.
Practices & PatternsVideo#community#video#benchmarks
Community
1 more matching entry in the full directory.
From the community
Posts from builders shipping with Jev right now.
Unclutter: an ad and slop blocker that runs on Jev
An open-source BS meter for debates and investor calls
🚨 Open Source Jev BS meter you can use this to analyze any debate / investor call / interview / sales pitch / podcast video fact check live , for example this dario interview cost 60 Jev calls / 111K tokens / $0.0047 github.com/ChetasLua/jevm…
🚨 I gave the Trump vs Kamala debate a live BS meter using Jev every sentence, both candidates, 5 yes/no questions each 1,191 Jev calls / 1.18M tokens / 415 ms median total cost : $0.0497 same questions for both, clips picked by one fixed rule, not a fact-check
A Jev-shaped model on Cerebras and Qwen
Built an alternative version of @typesafeai but on @cerebras with Qwen 3.8 27b. Similar quality, similar performance, but vastly different cost. TypeSafe was way cheaper, and did beat Qwen on performance. Closest we can get using LLMs I think. Source: github.com/iammrduncan/ty…
@CompleteSkeptic
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x
Jev benchmarked on two public safety corpora
1/ Benchmarked TypeSafe's Jev on two public safety corpora. It doesn't generate text, it returns calibrated probabilities you threshold in code. 96.5% on prompt injection, all 662 messages in deepset/prompt-injections. No tuning. 325ms p50.
@CompleteSkeptic
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x
An open 151M-parameter decision engine
TypeSafe AI came out of stealth with Jev, and access is behind a waitlist. I built an open source version Verdict (Open-jev) you can run right now in a browser tab: And its a real post trained model..(link in comments) It is a post trained 151M parameters model. ModernBERT-base Show more
@CompleteSkeptic
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x
Agent Handoff Gate: agents verify what they hand off
I built Agent Handoff Gate with typesafe Jev An experimental protocol for AI agents to verify worker evidence before handing results back to the lead. Less blind trust, fewer useless review loops. github.com/zsoXi/agent-ha…